© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128045
    Prev
    Next
    INDEX
    Explanations

    This neuron spots numeric literals (numbers, especially floats) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    ending sequence

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     bains
    -0.97
     demandes
    -0.89
    newValue
    -0.88
     kuchen
    -0.88
    uristic
    -0.87
    isław
    -0.87
     licha
    -0.85
     MARIE
    -0.85
    🏩
    -0.84
     unrival
    -0.83
    POSITIVE LOGITS
     with
    1.45
     även
    1.08
     With
    1.05
     Each
    1.00
     then
    0.98
     three
    0.93
     Then
    0.91
     mischievous
    0.91
     After
    0.91
     Even
    0.90
    Activations Density 0.010%

    No Known Activations